Wei Jiang

Results 3 issues of Wei Jiang

## Describe the bug 在使用`nn.CrossEntropyLoss()`时,指定`weight`参数为`jitter.ones(num_classes, dtype='float32')`和置为`None`的运行结果不同。且指定包括全一在内的值后会带来梯度爆炸。 ## Full Log 这是指定了`weight`具体值的训练曲线。 ![accuracy](https://user-images.githubusercontent.com/74121851/196984829-1bd2219b-e562-4ffa-9472-114318089e41.png) ![loss](https://user-images.githubusercontent.com/74121851/196985000-7867e00f-f93b-4bdf-bbac-d93285dbc51e.png) 这是不指定`weight`值的训练曲线。 ![accuracy](https://user-images.githubusercontent.com/74121851/196985175-7f1a28ac-0e35-48e6-abf3-3c812689fe22.png) ![loss](https://user-images.githubusercontent.com/74121851/196985198-def22249-a974-4b64-809f-370afc4a6070.png) ## Minimal Reproduce 数据集是CIFAR-10,简单的Conv+Linear,类似于Jittor教程的MNIST分类网络。 ## Expected behavior 希望两者训练行为至少一致,最好是都不会发生梯度爆炸。

How could I use the GRU Fusion module to feed-forwardly fuse some multiview pictures? The checkpoint `G.ckpt` seems to be somehow broken and the `representations.grufusion` is not even included in...

在点击课程后未能实现向后跳转,需修正。